Trends in forms of civic involvement in the Netherlands between 2008 and 2020

ABSTRACT

Dutch civil society is seen as well-equipped and known for its high level of civic involvement in various fields. For sustainability of civil society, however, it is crucial to understand to what extent civic involvement changes over time. Therefore, this article describes how civic involvement in the Netherlands has developed in the period 2008–2020. Using longitudinal high-quality survey data, we consider (contradictory or complementary) causes for trends on civic involvement during the period of investigation: (1) ongoing individualization; (2) traditionalization; (3) the role of major societal events that may periodically undermine or boost civil society. We find an overall trend of predominantly stable, yet slowly and statistically significant declining engagement levels that seem robust and relatively unaffected by societal events. Our results also reveal clearly fixed sequences of forms of civic involvement between distinguished (clusters of) organizations. The results in this article are most in line with individualization processes. However, given that the declines are slow and relatively unaffected by societal events, the findings could also be due to the slow process of cohort replacement.

KEYWORDS:

• Civil society
• civic involvement
• activist organizations
• leisure organizations
• interest organizations

Introduction

This article aims to show how civic involvement in the Netherlands has developed over the period 2008–2020. It provides evidence for changes in participation in civil society (Dekker, 2002). Civil society refers to non-governmental action by ordinary citizens, ‘self-responsible’ inhabitants expected to look after their own and others’ affairs, without public help (Dekker, 2019). It represents ‘the ideal of a “caring society” or a “big society”, in which people feel responsible for their fellow citizens in need and for their neighborhoods’ (Dekker, 2019, p. 77). For the sustainability of civil society, it is crucial to understand to what extent civic involvement changes over time. Historically, Dutch civil society has expanded in the twentieth century (Burger & Veldheer, 2001), being of considerable importance and size in various fields (Dekker, 2013; Habraken et al., 2013). In comparative perspective, the Netherlands has one of the highest scores on the Civil Society Index (CSI) (Salamon & Sokolowski, 2004). CSI assesses the societal impact of civil society and whether it serves societies’ common good. Given its importance, Dutch civil society is, therefore, seen as ‘well-equipped’ (Dekker, 2013, p. 18).

Different typologies of forms of civic organizations exist in the literature. Gordon and Babchuk (1959) introduced the distinction between instrumental organizations (directed towards social influence) and expressive organizations (directed towards their members). Others differentiate between Olsonian (with a focus on self-interest) and Putnamesque organizations (with a focus on the contribution to social trust) (Knack, 2003), or between formal (volunteering as productive activity) and informal volunteering (helping) (Wilson & Musick, 1997). In this contribution, we base our typologies on the goals of the organizations (Lelieveldt et al., 2007). In doing so, we follow Gesthuizen et al. (2013) and Van der Meer et al. (2009), who distinguished three types of civic organizations: activist organizations, leisure organizations and interest organizations. Activist organizations encompass humanitarian organizations and environmental organizations. These organizations primarily serve broader interests in society that in general do not directly benefit the (socio-economic) interests of their members (Lelieveldt et al., 2007). Leisure organizations – like sports clubs and cultural clubs – offer recreational and socializing activities to their members. Interest organizations – as trade unions, business organizations or consumer organizations – defend and represent socio-economic interests of their stakeholders. In this article, we present an overview of Dutch civic engagement in the period 2008–2020, for organizations within these three types of civic organizations, next to engagement in other organizations that respondents in our research could indicate to be involved in such as religious organizations, political parties and other types of organizations.

In the study of civic involvement, different forms of engagement are distinguished for which we can provide trends. These forms include organizational membership and donating money as well as participating actively in an organization or volunteering for an organization. Gesthuizen et al. (2013) showed, using cross-national data, that forms of involvement vary across activist-, leisure and interest organizations. With regard to activist organizations, donating money was the most popular form of involvement, followed by membership whereas volunteering and participating in activities for these organizations was least popular. This typical sequence in the popularity of forms of involvement was equivalently present in most European countries. In leisure – and interest organizations, donating money was least popular and membership was the most common involvement form. This typical order of forms was also equivalently present in most European countries.

The proposition based on the robust cross-national findings, that forms of involvement vary per type of organization is only limitedly studied. Previous contributions (cf. Gesthuizen et al., 2013; Pichler & Wallace, 2009; Van der Meer et al., 2009) used a cross-sectional design, drawing on data that were only available in a particular year in a large number of European countries. Considering the availability of recent, longitudinal high-quality data in the Netherlands, we have the possibility to test the longitudinal robustness of forms of Dutch civic involvement within organizations belonging to the types of civic organizations over more than the recent decade. In doing so, we shed light on fluctuations in these forms of civic involvement over a period of twelve years. This enables us to inform on the sustainability of Dutch civil society and the cohesiveness of Dutch society. Moreover, this study provides a contribution by mapping trends in civic engagement for so many types of organizations over a long time period. As far as we know, such an encompassing empirical overview of historical trends in the landscape of civil society fulfils a lacuna in our knowledge.

We explore to what extent several societal trends, such as individualization, traditionalization versus period effects, such as the economic crisis of 2008, the refugee crisis of 2015 and the recent COVID crisis, are related to changes in the Dutch civic landscape. Originally put forward by Mannheim (1967), period effects concern exposure to societal conditions that equally affect all people in a society at a particular period in time simultaneously and not only people in their formative years; whereas cohort effects propose that exposure to the societal context during peoples’ formative years shapes the worldview of cohorts in an enduring way, distinct from other cohorts. Given the expected impactful societal conditions that occurred during our period of investigation, we explore fluctuations in forms of Dutch civic involvement. Hence, the research question is: To what extent have forms of Dutch civic involvement in different organizations changed in the years 2008–2020 and have these changes possibly been due to period effects?

Contradictory and Complementary Trends

We consider three contradictory and complementary perspectives that provide expectations on differential trends in civic involvement. The first is on the role of progressing individualization. Individualization theory contains the proposition that human identity is transformed from a ‘given’ into a ‘task’ (Bauman, 2001). Individuals are variously charged with the responsibility for performing that task and for the consequences of their performance (Beck, 1992; Beck & Beck-Gernsheim, 2002). Individualization, to quote Beck and Beck-Gernsheim (2002), creates ‘internalized freedom’ (p. 157), that may alienate many people from traditional institutions in society. Following the individualization thesis, individuals’ autonomy thus stands central. Although different in its theoretical analyses, the influential study of Putnam (2000) reaches similar conclusions. People would be increasingly likely to be dissociated from organizations, as typically summarized with the ‘bowling alone’ title of Putnam’s seminal work. This fits the process of secularization, with fewer church members present in religious organizations to help out with voluntary services (Van Ingen & Dekker, 2011a). There is more preliminary evidence for a decrease in (people who), respectively, donate money (Bekkers et al., 2017), volunteer (Damian, 2019), spend hours on volunteering (Bekkers & Ruiter, 2009) or have (an) organizational membership(s) (Kuyper et al., 2019) in the Netherlands. Also, the recent longitudinal study of Statistics Netherlands (2022) found a slight decrease in the number of people who volunteer. However, some of these contributions did not address a multiplicity of organizations, nor considered all forms of involvement in a longitudinal perspective. Following the individualization thesis, we derive the expectation that processes of individualization may gradually and slowly erode civil society and involvement forms.

The opposite second perspective on trends does not dismiss the process of individualization, but argues that it comes along with a counter-reaction of a willingness to keep certain traditions or to reintroduce ‘old’ traditions in society. From such a traditionalization perspective it is expected that after a period of erosion of civil society due to individualization, civil society is re-institutionalized by a part of society. Individualization and traditionalization can therefore coexist (Heelas, 1996; Hoffmann & Miller, 1997). Following this perspective, we may derive the expectation of stability in civil society and forms of involvement.

Empirical findings more in line with this second expectation are from Van Ingen (2008) who found no trend in formal involvement in associations between 1975 and 2000.¹ In addition, with different data, Van Ingen and Dekker (2011b) studied associational involvement between 1975 and 2005 and found no erosion of civil society. Also, Bekkers and Ruiter found that the percentage of volunteers had hardly declined in the 1990s. In turn, for the years 2001 to 2009, Schmeets (2013) even showed a slight increase in people ‘who currently volunteer’. And, notably, the data from Statistics Netherlands analysed by Arends and Schmeets (2018) show no decline in volunteering.

A third perspective on trends refers to the role of major societal events that may periodically affect all people in society similarly and hence undermine or boost civil society. Examples in our period of investigation are the economic crisis of 2008, the refugee crisis of 2015 and the recent COVID crisis that took off in 2020. Such major events likely supposedly impacted many members of civil society because it increased the salience of certain issues or restricted people to be involved in organizations. For example, the refugee crisis may have encouraged inhabitants in refugee receiving countries to become active in activist organizations (Pries, 2019; Van der Veer, 2022). Vice versa, changes in ethnic diversity may have undermined civil society, at least periodically, which is a proposition that may be derived from Putnam (2007). Therefore, we derive the expectation that these large periodic societal events induce versus reduce civil society, i.e., all forms of involvement.

Methods

Data

The archived and open-access data were derived from the Longitudinal Internet Studies for the Social Sciences (LISS), a household panel that is aimed to be representative of the population of the Netherlands of 16 years of age and older.² LISS provides high-quality survey data, and is developed to monitor changes in the life course and living conditions of the panel members aimed to represent the general Dutch population (Scherpenzeel, 2009). The panel is a random sample of Dutch household addresses, drawn by Statistics Netherlands. After the selected households are invited and informed with a letter, a face-to-face recruitment interview is conducted with them. When their response is positive, every household member with a minimum age of 16 years is able to participate in the panel. Self-selection into the panel is hence not possible.

This recruitment method to refresh the panel is repeated every two to three years, starting from 2007. To measure the degree of (selective) attrition and panel representativeness, the dataset is compared to the population figures of Statistics Netherlands. This is monitored throughout the year for specific socio-demographic characteristics on the household level, and for individual members of the household.³ When there is dropout and the panel becomes too small, a new recruitment round is started to reach the desired number of at least 5,000 respondents. A new recruitment round is also launched when there arises a threat of significant deviation from population figures of Statistics Netherlands, on specific characteristics as mentioned in note two. When attrition is selective on specific characteristics, Statistic Netherlands starts a stratified sampling round in order to overcome sampling error in the LISS panel.⁴

Data are annually collected in two fieldwork periods of both three to four weeks. A reminder was sent twice to non-responders. A second fieldwork period is directed to those who did not respond in the first fieldwork period, again followed by two reminders. Questionnaires are filled out online. The survey includes non-internet users as well. The data had a minimum of 5,051 respondents in 2019 and a maximum of 7,352 respondents in 2008. There are three reasons as to why the number of observations varies between the years. One, the selected number of household members who are part of the LISS panel differs from year to year, due to attrition and refreshment samples. Two, the annual data collection can be planned close to or close after a necessary recruitment round (see note four for an overview). Three, the respondents’ response rate differs from year to year.⁵

The yearly retention rate is about 90% and, as said, refreshment samples are drawn aimed to ensure panel representativeness. LISS delivers their data, however, unweighted. To ensure representativity, the authors additionally constructed and used a weight variable on the basis of gender and age.⁶

We make use of the data of the annual module on Social Integration and Leisure, that was collected between the years 2008 and 2020. We also use the background variables module, because it contained the necessary variables to construct the weight variable. This is a monthly module in which, for the years 2008 to 2020, we have selected the December module each time. Finally, datasets of all included years were merged into one dataset.

Measurements

Dependent Variables

Respondents were asked about forms of civic engagement by answering the question: ‘There are a number of organizations that you are free to join. Can you indicate, for each of the organizations listed, what applies to you at this moment or has applied to you over the past 12 months?’ We distinguish between all the answer categories of the forms of involvement: (1) member; (2) donated money; (3) participated in an activity; (4) performed voluntary work. People not involved in any of these forms indicated to have ‘no connection’. Selected organizations in which respondents could perform their form of civic involvement are: (1) organizations for humanitarian aid, human rights, minorities or migrants; (2) organizations for environmental protection, peace organizations or animal rights organizations; (3) sports clubs or clubs for outdoor activities; (4) cultural organizations or hobby clubs; (5) trade unions; (6) business, professional or agrarian organizations; (7) consumers’ organizations or automobile clubs; (8) religious or church organizations; (9) political parties; (10) science, education, teachers’ or parents’ organizations; (11) youth, pensioners, women or friends’ clubs; (12) other organizations. Note that LISS, in the wave of 2020, divided organizations for humanitarian aid, human rights, minorities or migrants into, on the one hand: ‘organizations for humanitarian aid and human rights’ and, on the other hand: ‘organizations for migrants’. We, however, combined these organizations to equal the measurement in the years 2008–2019. In the second stage of the analyses, similar types of organizations are clustered. This indicates that humanitarian and environmental organizations represent the category of activist organizations. Next, sports clubs or clubs for outdoor activities and cultural organizations or hobby clubs constitute leisure organizations. Subsequently, trade unions, business, professional or agrarian organizations and consumers’ organizations or automobile clubs represent interest organizations. Remaining organizations (see (8) religious organizations to (12) other organizations here above) are not clustered, since there is no convincing argument in the theoretical line of reasoning we follow, that these organizations belong to one similar cluster. Therefore, these organizations are left out in the second stage of the analyses.

Independent Variable

To measure the influence of period effects on trends in forms of involvement, year was measured by including all years of investigation. Years 2008 to 2020 were then recoded into 0 (2008) running to 12 (2020). Following Van Ingen (2008), period effects are, thus, simply captured by years of measurement.

Methods

We show trends in the incidence of the forms of involvement in all twelve distinguished organizations in the period 2008–2020. Note that the percentages on the left axis might differ per figure. When this is the case, it is done for readability. For the most popular form per organization we included a trend line, showing the estimated course of this particular form of involvement. Next, we calculate the percentage of change over the years for the two most popular engagement forms per organization. These calculations give further understanding of how sizeable changes in the trends are.⁷

Next, we test whether (changes in) trends in forms of involvement are significant over time. Using regression analysis, we estimate a model to test whether the trend in each form for each organization is statistically significant. In total, 48 models were estimated. The results of the regression analyses can be found in Table 1.

Table 1. Regression Parameters of Years on Different Engagement Forms in Different Types of Organizations.

	Membership		Donating		Participation		Volunteering
	b	S.E.	b	S.E.	b	S.E.	b	S.E.
Humanitarian Organizations	−0.161***	0.034	−0.589*	0.216	0.035	0.021	−0.011	0.021
Environmental Organizations	−0.296***	0.035	−0.409*	0.134	0.053**	0.014	0.018*	0.008
Sports / Outdoor Clubs	−0.260**	0.061	−0.048*	0.020	−0.339***	0.070	−0.136*	0.047
Cultural / Hobby Clubs	−0.334***	0.038	−0.058	0.027	−0.241**	0.065	−0.079*	0.030
Trade Unions	−0.334***	0.037	−0.018	0.011	0.001	0.011	−0.008	0.007
Business / Professional / Agr.	−0.295**	0.069	−0.019*	0.007	−0.131**	0.035	−0.031*	0.013
Consumer / Automobile Clubs	−0.550***	0.074	−0.020	0.011	−0.045*	0.015	−0.011	0.009
Religious Organizations	−0.449***	0.045	−0.529***	0.076	−0.173***	0.031	−0.162***	0.011
Political Parties	−0.079**	0.023	−0.046*	0.018	0.007	0.016	−0.002	0.009
Science / Education etc.	−0.246***	0.034	−0.063*	0.025	−0.179*	0.068	−0.123**	0.032
Youth / Pensioners etc.	−0.226***	0.043	−0.018	0.010	−0.069	0.041	−0.099**	0.028
Other Organizations	−0.293***	0.033	−0.103*	0.039	−0.086**	0.023	−0.141***	0.021

*p < 0.05, **p < 0.01, ***p < 0.001.

Source: LISS (2008–2020).

In the second stage of the analyses, we present the trend in involvement in the three clusters of activist, leisure and interest organizations. Again, for the most popular form of involvement within the cluster, we estimated a trend line.⁸ Similarly, regression analyses have been performed to test whether (changes in) trends in forms of involvement within a cluster are significant over time. The results of these analyses can be found in Appendix 1.

Results

Trends in Involvement in Humanitarian and Environmental Organizations

We present an overview of trends in forms of civic involvement for the various voluntary organizations over the period 2008–2020, starting with the two organizations that constitute activist organizations. Figure 1 shows the longitudinal trends in forms of civic involvement for humanitarian organizations. Next, Figure 2 reveals trends for environmental organizations. From both figures it is clear that donations are the most popular form of involvement, showing a slow and steady decrease to 2019. However, and remarkable, the number of people giving donations increased in the COVID year 2020. The year 2020, therefore, deviates from both donation trend lines in the figures.

Figure 1. Longitudinal trends in forms of civic involvement in humanitarian organizations.

Figure 2. Longitudinal trends in forms of civic involvement in environmental organizations.

When looking at the size of the changes in donations to humanitarian organizations, it becomes clear that the relative decrease compared to the starting year 2008 is indeed larger in 2019 (−39%) than in 2020 (−11%). This corresponds to an annual decrease of 3% (2019) and less than 1% (2020). The same applies to environmental organizations, where since 2008, the decrease in 2019 was – 35% compared to – 17% in 2020. Annually, this decrease is 2.7% (2019) and 1.3% (2020).

Membership is the second most popular form of involvement in both organizations. The pattern over the years seems rather stable, however, the relative change in memberships is a sizeable decline of – 19% for humanitarian organizations (−1.5% annually) and – 28% for environmental organizations (−2.2% annually), compared to 2008. Volunteering and participating in activities are the least popular involvement forms. Despite their unpopularity, the involvement patterns for humanitarian and environmental organizations seem rather stable between 2008 and 2020. During the period of investigation, involvement patterns of humanitarian and environmental organizations show a fixed order of (un)popularity.

Trends in Involvement in Activist Organizations

Figure 3 presents patterns of involvement over time for the whole cluster of activist organizations. This thus constitutes the combined longitudinal trends from humanitarian and environmental organizations, accounting for the possibility that people are active in both humanitarian and environmental organizations.⁹ Mirroring the findings of engagement in the separate organizations, donations are most popular and the course of the grey line that belongs to donations is very similar to that of the individual organizations as depicted in Figures 1 and 2. The decrease is in particular visible between 2011 and 2019. Therefore, the sudden increase in donations in 2020 does not fit the prediction of the trend line for the year 2020. The size of the changes in donations over time are substantial for the period 2008–2019 (−36%) and smaller if the sudden increase in 2020 is captured (−13%). For memberships there is a relative decrease of minus 21% (and about minus 1.7% annually). Following the patterns of humanitarian and environmental organizations, volunteering and participating in activist organizations are least popular for the cluster of activist organizations. The trends are, however, fairly stable over time.

Figure 3. Longitudinal trends in forms of civic involvement in activist organizations.

Trends in Involvement in Sports and Outdoor Clubs and Cultural Organizations or Hobby Clubs

Figure 4 outlines longitudinal trends for sports and outdoor clubs and Figure 5 does so for cultural organizations or hobby clubs. For both organizations, memberships are the most popular form of civic involvement. In the case of sports and outdoor clubs, memberships are even by far the most popular. In both organizations, the trend line does barely deviate from the actual membership line, showing a very stable and gradually declining course over the years. Relatively, mirrored against the year 2008, this declining course in memberships is reflected in a decrease of – 13% for sports and outdoor clubs (−1% annually) and – 23% for cultural organizations or hobby clubs (−1.8% annually).

Figure 4. Longitudinal trends in forms of civic involvement in sports and outdoor clubs.

Figure 5. Longitudinal trends in forms of civic involvement in cultural and hobby clubs.

Participating in activities is the second popular involvement form for both organizations. Relative changes over time are, nevertheless, sizeable. Compared to 2008, about 28% fewer people participated in activities for sports and outdoor clubs. This number is – 25% for cultural organizations or hobby clubs. Volunteering and donating money are least popular. These two involvement forms picture a rather stable course from 2008 to 2020. The order of (un)popularity in forms of involvement is the same for sports and outdoor clubs and cultural organizations or hobby clubs, as well as that the order of the forms of involvement is fixed for both organizations over the full period of investigation.

Trends in Involvement in Leisure Organizations

Figure 6 presents trends for leisure organizations, which is formed out of the combination in sports and outdoor clubs and cultural or hobby clubs. Not surprisingly, given the trend as shown in especially Figure 4, memberships are by far most popular. The pattern, with an almost identical course as the trend line, displays a slow gradual decline over time. Relatively, it comes down to a decrease of 11% over the period 2008–2020, so a decline of less than 1% per year. Participation in activities is the second popular form of engagement for leisure organizations. Figure 6 depicts a decrease from 2008 to 2011 and stability to 2017, followed by a decline in 2018 and 2019 and some recovery in 2020. Despite the increase in 2020, the relative size of the changes is still a substantial minus 24% (and a – 1.8% per year). Volunteering for and donating to leisure organizations are least popular. Both patterns seem fairly stable between 2008 and 2020, although there seems to be some tendency towards less volunteering.

Figure 6. Longitudinal trends in forms of civic involvement in leisure organizations.

Trends in Involvement in Trade Unions, Business Organizations and Consumer Organizations

Figures 7–9 show longitudinal involvement patterns for, respectively, trade unions, business, professional or agrarian organizations and consumers’ organizations or automobile clubs. Especially for trade unions and consumers’ organizations or automobile clubs memberships are the most (and almost only) popular involvement form. Over the years, the trend line from memberships for all of these three organizations reveals a stable, slowly decreasing trend toward less memberships. We only see deviations from the trend line as of the year 2018 with respect to business organizations.

Figure 7. Longitudinal trends in forms of civic involvement in trade unions.

Figure 8. Longitudinal trends in forms of civic involvement in business, professional and agrarian organizations.

Figure 9. Longitudinal trends in forms of civic involvement in consumer and automobile clubs.

The finding of decreasing trends is supported by the relative changes over time, given the decrease in memberships with regard to trade unions (−19% and – 1.4% yearly) and consumers’ organizations or automobile clubs (−24% and annually – 1.9%). Business organizations even note a decrease in memberships of – 47% in the period 2008–2018 (−3.6% per year), a sizeable decline that recovers somewhat in 2020.

Participating in activities is the second popular involvement form. However, it is rather unpopular for especially trade unions and consumers’ organizations, while volunteering and donating money are in general unpopular for all involved organizations. We, therefore, only picture the relative decline in participation in activities for business organizations, which is a substantial minus 29% (−2.2% annually). Lastly, all forms of involvement within the three organizations reveal a fixed order over the full period of investigation.

Trends in Involvement in Interest Organizations

Figure 10 presents trends in forms of involvement for the cluster of interest organizations, constituted by trade unions, business organizations and consumers’ organizations. The figure shows that interest organizations rely on their members. The blue membership line shows a declining course, with some slight recovery as of 2018.

Figure 10. Longitudinal trends in forms of civic involvement in interest organizations.

Participation in activities is the second popular engagement form. It jumps out slightly in popularity compared to donating and volunteering, because of the involvement patterns from business organizations as shown in Figure 8. A slight decrease in participation is visible over the years. This is reflected in the relative changes over time, with a sizeable decrease of – 35% in 2008–2020 and an annual decrease of – 2.7%.

Volunteering and donating to interest organizations are rather unpopular. Their patterns seem fairly stable over time, with few people who are involved in interest organizations via these ways.

Other Organizations

Figures 11–15 contain the longitudinal trends in forms of involvement for other civic organizations. All taken together, memberships are the most popular involvement form. The trend line of memberships reveals slightly downward going trends from 2008 to 2020 for all organizations. When looking at the size of the changes, we note that they are substantial e.g., for political organizations where the decrease is 17% since 2008 (−1.3% yearly). Another finding that is noteworthy, is the relative and sizeable decrease in donations to religious and church organizations. This declined by half in the period 2008–2018, recovering slightly to a decrease of one-third in the period 2008–2020.

Figure 11. Longitudinal trends in forms of civic involvement in religious and church organizations.

Figure 12. Longitudinal trends in forms of civic involvement in political parties.

Figure 13. Longitudinal trends in forms of civic involvement in scientific, education, teachers’ and parents’ organizations.

Figure 14. Longitudinal trends in forms of civic involvement in youth, pensioners, women and friends’ clubs.

Figure 15. Longitudinal trends in forms of civic involvement in other organizations.

Statistical Analyses of Trends in Forms of Civic Involvement

Table 1 shows the regression parameters of time (as measured by year of study) on the different engagement forms for the different types of organizations. When looking at the table, it becomes clear that memberships show an overall statistically significant decrease for all kinds of organizations. Donations show also an overall decrease, however, only significant for activist organizations, (some) leisure organizations and (only one type of) interest organization. Participation shows a statistically significant increase for (at least one) activist organization, however, significant decreases for leisure and interest organizations. Volunteering is rather stable in activist organizations, however, shows a significant decrease in leisure and (at least one) interest organization.

Discussion

Dutch civil society is well-known for its size and civic involvement in various domains. It is, however, vital for the sustainability of civil society to get an understanding of how civic involvement changes over the years. Therefore, this article addressed how (contradictory and complementary) perspectives on how trends in civic involvement would develop, were empirically supported in the Netherlands, between 2008 and 2020. An encompassing overview of historical trends in the Dutch civic landscape for so many different organizations fills a scientific lacuna, not only in the Netherlands but is also missing in many other countries. We used longitudinal high-quality survey data and considered three perspectives on why there would be differential (contradictory and complementary) trends in civic involvement in recent times, with (1) individualization; (2) traditionalization; (3) period effects of major societal events that might undermine or boost civil society, such as the economic crisis, refugee crisis and the COVID crisis that all occurred in the period under study.

We discovered an overall trend of predominantly stable, yet slowly and statistically significant declining engagement levels that seem robust and relatively unaffected by societal events. This indicates that neither the economic crisis, nor the refugee crisis or (part of) the COVID crisis substantially changed forms of civic involvement in Dutch civil society. As an exception to this general pattern of findings, we may list the increase of donations to activist organizations in the COVID year 2020. However, overall, we find no support for the perspective about period effects of major societal events that might undermine or boost civil society. Instead, the findings of slow and small, yet significant decreases in most forms of involvement is most in line with the perspective of the individualization processes that erode civil society. We thus find encompassing support for what Bekkers et al. (2017), Damian (2019), Statistics Netherlands (2022) and Kuyper et al. (2019) found in studies focusing on single organizations or single forms of involvement. The longitudinal decrease in civic engagement indicates that we barely find support for the second perspective about traditionalization that proposes stability or even a re-institutionalization of civic involvement.

Next to the decrease in civic engagement, results showed clearly fixed sequences of popularity of forms of civic involvement between (clusters of) activist-, leisure and interest organizations. These results are in line with findings by Gesthuizen et al. (2013). However, their contribution had a cross-national design, whereas we pictured the ‘fixedness’ of this demarcation by observing it for a period of twelve years in Dutch civil society. These fixed orders were similar for organizations belonging to the same clusters, respectively, activist-, leisure and interest organizations. Next, the fixed rank order suggests that memberships are the most common form of engagement, while donations and participating in activities are considerable less popular, and volunteering the least common. One reason for this order might be that organizations recruit donors and volunteers from their ranks of members. Several contributions, however, suggest that this regularity may have declined over time (e.g., Cnaan & Handy, 2005; Lee, 2020; Macduff, 2005). This claim is not supported in this study (see Appendix 2). Trends are rather stable indicating that a stable percentage of the members volunteers, with an indication of a slightly larger share of members volunteering for activist organizations. We should note, however, that memberships are not the most popular engagement form in activist organizations.

A suggestion that has been made in earlier studies is that society has turned to ‘checkbook participation’ (as argued by Stolle et al., 2005), since memberships of more ‘distant organizations’ (such as humanitarian organizations, trade unions or consumers’ organizations) have decreased strongly. However, we find no support for that idea since also for donating money, the general pattern is a small decrease (or statistical insignificance), and not an increase. As such, the prediction of small annual rises in donations up to 2015 as made by Bekkers and Ruiter (2009) is not supported in this study.¹⁰ Given the fact that the declines in engagement levels are relatively slow and relatively unaffected by periodic societal events, it is also likely that cohort effects are (partly) responsible for the observed trends; older more involved citizens would become replaced then by younger less actively engaged citizens. These younger cohorts may be active in new domains of civic engagement, such as posting messages on issues addressed by the organizations on social media, but they would then be less likely to engage in the forms of involvement in organizations we studied here. An alternative explanation for the declines we found might be that organizations have changed and have invested less in mobilizing engagement or have become less effective in these efforts. Bekkers and Van Teunenbroek (2020) provide some evidence for this.

In the midst of the slowly declining civic engagement levels, humanitarian and environmental organizations had a rise in the share of people who donated money to them in 2020. This increase in donations is remarkable, because the COVID restrictions impeded going from door to door asking for donations. It is also remarkable when the refugee crisis is considered, which caused no increase in donations given to humanitarian organizations. Possibly, people felt more inclined to spend their (saved) money on organizations that support society when they experience a crisis that affects them personally. Moreover, it is relatively easy to ‘just wave with the checkbook’ as an example of checkbook participation (Stolle et al., 2005) when people feel more inclined to prosocial behaviour. In addition, it might be that activist organizations received their donations online during the COVID restrictions. Online donations and crowdfunding are becoming more popular for civic organizations (see for instance De Wit & Bekkers, 2020; Van Teunenbroek & Bekkers, 2019), which might partly explain the rise in people who donated money to humanitarian and environmental organizations in 2020.

A limitation of this study is that the questionnaire does not include an item covering involvement in health organizations. For donations, this is the most common area in which the Dutch population gives money to nonprofit organizations. A study on donations in the Netherlands (Bekkers et al., 2020) shows that two thirds of households in the Netherlands give to health causes. In addition, that study (Bekkers et al., 2022) reveals that Dutch donate most to health causes in the Netherlands. This implies that the overall level of donations cannot be retrieved from our study; we only provide the trends for numerous organizations that have been included.

Relevant is how the overall trend of gradually but significantly declining involvement levels in Dutch civil society will evolve after 2020. And how the COVID crisis (aftermath) impacts on (Dutch) civil society. Based on our findings we expect a continuation of the slowly decreasing engagement levels. Given that we showed the absence of period effects of major societal events, we expect that the COVID crisis will not disrupt this pattern. More challenges for civic organizations arise, however. The current high inflation and energy prices may make people want to save money on memberships or donations. This could be a potential risk for organizations that rely heavily on these forms of engagement. Moreover, there is an ongoing shortage in personnel in many countries’ labour market. That may come at the cost of people willing to volunteer. The coming years should reveal whether these processes affected Dutch civic engagement in some way.

This study shows that Dutch civic landscape is slowly yet significantly eroding over time, but is, however, persistent against societal shocks and (major) events. In the middle of this civic erosion, memberships seem the main form of concern, as all (clusters of) organizations noted a decline in amount of members over the years. This is especially critical for organizations that have memberships as most popular involvement form, such as leisure and interest organizations. Civil society organizations, therefore, are challenged to invent strategies and policies that focus on binding their members and recruiting new ones. One positive exception in the downward trends is the participation in and volunteering for environmental organizations, which have gained significant popularity both in participation as well as in volunteering, mirroring the growing awareness and urgency to address environmental issues. But overall, this study provides evidence for a slowly but steadily decline in civic involvement.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Sustainable Cooperation - Roadmaps to Resilient Societies (SCOOP) [grant number 024.003.025.].

Notes

1 Instead, Van Ingen (2008) found a cohort effect: younger cohorts showed less activity in formal participation between 1975 and 2000.

2 The data can be found on https://doi.org/10.17026/dans-zaf-casa.

3 These specific socio-demographic characteristics are: gender, age, educational level, net income per household, urbanity, province, household size and residential form. Note that this information is obtained via personal communication with an associate from LISS.

4 The following recruitment rounds have been done in the past: initial random sampling round (2007); stratified sampling round (2009); random sampling round (2011); stratified sampling round (2013); stratified sampling round (2016); stratified sampling round (2019) and random sampling round with a targeted approach (2022). Additional documentation from LISS shows that, as of 2015, lower educated, young people and lower incomes are somewhat underrepresented. In turn, higher educated and people older than 65 or pensioners are somewhat overrepresented. As said, the recruitment rounds try to correct this. Note that this information is obtained via personal communication with an associate from LISS.

5 See for the year 2008: https://dataarchive.lissdata.nl/study_units/view/7 (then click on Response Information). See for the year 2019: https://dataarchive.lissdata.nl/study_units/view/993 (then click on Response Information).

6 In creating the weight variable, the authors obtained population numbers from Statistics Netherlands on the basis of gender and age for the years 2008 to 2020. Age was divided in four categories: 16 to 29, 30 to 49, 50 to 64 and 65+. This led, consequently, to 8 combinations (men in the four age categories and women in the four age categories) for every individual year. Next, the division in the LISS data was obtained for all involved years on the basis of gender and the four relevant age categories. Then, a weight was constructed to correct for over – or underrepresentation of gender and age in the LISS data. This weight has been applied in our analyses. We acknowledge that we also preferred to include educational level in the weight variable (as there is more selective attrition over time, but also stronger repairs when replenishing the sample), however population data on the basis of educational level is lacking, as population data of educational level derived from Statistics Netherlands is only an approximation. Note that the unweighted data generated almost the same results as the weighted data. For instance, with regard to Table 1, there is only a difference in donations for political parties (which turned significantly negative in the weighted approach). The results from the unweighted approach are available upon request by the first author.

7 All calculations for all involvement forms per organization are available upon request by the first author.

8 All calculations for all involvement forms per cluster are available upon request by the first author.

9 Percentages in the figures and relationships in the table in Appendix 1 from the cluster of organizations might, logically, differ from the results from the ‘separated’, individual organizations. One reason might be that respondents can be, for instance, a member of both a humanitarian and an environmental organization (thus score two times a ‘1’ on this item in the separated results), while they, in this case, only score a single ‘1’ on membership of an activist organization. So, it is not the case that two memberships of individual organizations that together form a cluster also count twice in the clustered results. It is then considered a single membership.

10 Fifteen separate regression analyses of all organizations and clusters of organizations over the period 2008-2015 have been executed. Thirteen of them noted a significant decrease in donations between 2008 and 2015. Two regression analyses showed insignificant results, although both directions of the unstandardized coefficient were negative. In sum, the results were as follows: (1) humanitarian organizations b – 1.210, p < .01; (2) environmental organizations b – 0.782, p < .01; (3) activist organizations (cluster) b – 1.335, p < .001; (4) sports organizations b – 0.103, p < .01; (5) cultural organizations b – 0.121, p < .01; (6) leisure organizations (cluster) b – 0.183, p < .01; (7) trade unions b – 0.054, p < .05; (8) business organizations b – 0.035, p < .01; (9) consumers’ organizations b – 0.045, p > .05; (10) interest organizations (cluster) b – 0.108, p < .05; (11) religious organizations b – 0.736, p < .001; (12) political parties b – 0.127, p < .001; (13) science organizations b – 0.142, p < .05; (14) youth organizations b – 0.013, p > .05; (15) other organizations b – 0.190, p < .01.

Appendices

Table A1. Regression parameters of years on different engagement forms in different clusters of organizations.

	Membership		Donating		Participation		Volunteering
	b	S.E.	b	S.E.	b	S.E.	b	S.E.
Activist Organizations	−0.314***	0.048	−0.678*	0.227	0.069*	0.028	0.009	0.023
Leisure Organizations	−0.368***	0.054	−0.081	0.039	−0.465***	0.103	−0.190**	0.052
Interest Organizations	−0.771***	0.082	−0.053*	0.019	−0.205**	0.049	−0.066**	0.015

*p < 0.05, **p < 0.01, ***p < 0.001.

Source: LISS (2008–2020).

Table A2. Percentages of members of cluster of organizations that also volunteer.

	Activist organizations	Leisure organizations	Interest organizations
2008	8.0%	18.8%	3.5%
2009	7.5%	16.4%	2.9%
2010	6.5%	15.2%	2.2%
2011	9.6%	16.7%	2.7%
2012	8.9%	15.1%	2.7%
2013	8.8%	16.8%	2.4%
2014	10.0%	15.8%	2.5%
2015	11.0%	15.4%	3.0%
2016	10.1%	17.3%	2.8%
2017	11.9%	15.4%	2.6%
2018	8.8%	15.1%	2.7%
2019	10.4%	16.3%	2.4%
2020	10.0%	17.8%	2.7%

Source: LISS (2008–2020).